A synthetic intermodal scene combines a train at left, container truck in front, ship and cranes at right, and a jet overhead.
Illustrative freight modes, not evidence of a real shared lane or a measured modal comparison.
Diagram text description

The inspected synthetic scene shows a train with container wagons on the left, a truck in the centre, a container ship with port cranes at right, and an airborne jet above. Containers, lights and wet pavement appear at dusk or dawn. No clearly readable carrier, cargo mass, timetable or emission value is established. It does not prove a freight-only aircraft, available services, common endpoints or an actual shipment result.

Comparing freight emissions by mode is a purchasing decision about a delivery, not just a contest between aircraft, trucks, trains and ships. First establish what must arrive, where and by when. Then compare the complete services that can perform that task, including their connectors and handling. A low mainline emissions figure cannot compensate for a missed deadline or an omitted delivery leg.

Broad mode averages can help identify alternatives worth investigating. They do not tell you the footprint of a particular consignment, the availability of a train departure or whether a vessel reaches the required port. Service details, loading, allocation and the emissions boundary determine whether the figures answer the same question.

The approach below uses one fictional cargo task throughout, a dimensioned rail-chain calculation and a four-offer comparison. All example schedules, coefficients and emissions figures are invented teaching inputs, not carrier quotations or measured savings. The aim is to make a defensible selection question visible before requesting real offers.

Start with the same transport task

Give every provider the same service specification. In the fictional comparison below, changing mode is allowed; changing the cargo, receiving site or required arrival is not. Distances need not be identical because each network takes a different route between the same endpoints. Requiring identical kilometres would disguise that operational difference rather than make the services equivalent.

Fixed itemCommon specification for every fictional offer
Cargo and condition2 tonnes of boxed, non-hazardous spare parts; 4 cubic metres; no refrigeration
Mass basisIncludes shipper-provided packaging, not carrier equipment or container tare
EndpointsCollection at Factory A and delivery to Buyer B, not terminal-to-terminal substitution
Arrival requirementGoods available at collection; delivery within 72 elapsed hours
Included serviceAll required pickup, mainline, transfers, handling and final delivery

Table 1, shown one row at a time. Every field is retained.

Row 1

Fixed item
Cargo and condition
Common specification for every fictional offer
2 tonnes of boxed, non-hazardous spare parts; 4 cubic metres; no refrigeration

Row 2

Fixed item
Mass basis
Common specification for every fictional offer
Includes shipper-provided packaging, not carrier equipment or container tare

Row 3

Fixed item
Endpoints
Common specification for every fictional offer
Collection at Factory A and delivery to Buyer B, not terminal-to-terminal substitution

Row 4

Fixed item
Arrival requirement
Common specification for every fictional offer
Goods available at collection; delivery within 72 elapsed hours

Row 5

Fixed item
Included service
Common specification for every fictional offer
All required pickup, mainline, transfers, handling and final delivery

Capacity, cargo acceptance and handling protection also have to fit. A service that cannot carry the volume or maintain the required condition is not rescued by a favourable carbon estimate. Record collection cut-offs and promised arrival at the actual buyer, not just arrival of the main vehicle at a hub.

These are fictional endpoints in a corridor assumed to have road, rail, air and coastal shipping access. They are not proof that any real lane offers all four choices. Use named sites and bookable services in a real enquiry. If a provider proposes a different delivery window, retain it as a separately labelled scenario rather than quietly changing the original test.

Use the same unit and boundary

A tonne-kilometre represents one tonne of freight moved one kilometre. For a leg calculated from mass and distance, multiply cargo tonnes by the distance basis required by the method, then by an intensity with matching units. Do this for the shipment's allocated share; a whole vehicle's emissions are not automatically its customer's result.

For example, take the fictional 2-tonne cargo over a 50-kilometre pickup leg. Its activity is 2 tonnes × 50 km = 100 tonne-kilometres. An invented coefficient of 0.10 kg CO2e per tonne-kilometre gives 100 tonne-kilometres × 0.10 kg CO2e/tonne-kilometre = 10 kg CO2e. The activity units cancel. This coefficient is only for the teaching case, not a recommended road factor.

Keep vehicle-kilometres, passenger-kilometres and freight tonne-kilometres distinct. Likewise, a price per shipment is not physical transport activity. Check whether distance is shortest feasible, great-circle or another method-compatible basis; do not pair a factor with a different distance convention without the required documented treatment.

The emissions numerator must match too. Operational-only CO2 and fuel/energy-supply-inclusive CO2e are not equivalent boundaries. State the gases, warming convention, upstream energy coverage and included hubs. If one offer excludes relevant stages, obtain the missing compatible information or identify the comparison as incomplete. A generic correction percentage would conceal the very difference the buyer needs to understand.

Air freight: urgency changes the decision

Air can be useful where the required speed genuinely constrains the shipment. Broad comparisons often show high air-freight intensity, but neither a universal multiplier nor a blanket ban answers a specific delivery problem. Examine the full collection-to-delivery promise, aircraft service, transfers and allocation rather than treating flight time as the whole journey.

Suppose an offer delivers the fictional cargo in 18 hours, while another eligible service promises 50 hours against the same 72-hour deadline. Air's extra speed is not required by that stated task. If a critical component instead must arrive within 24 hours, the task has changed: reassess the feasible set rather than pretending the slower offer still meets it. For real air results, distinguish freighter from passenger-aircraft cargo allocation and include airport access and final road delivery. A stop through an intermediate hub also changes the chain. Where only part of an order is urgent, a split shipment may be worth investigating, but its two complete services must still satisfy the actual requirements; it is not an automatic reduction claim.

Road freight: vehicle and load matter

Road's flexibility is valuable when it connects the exact collection and receiving sites. Its emissions depend on vehicle class, energy source, loading, empty operations and route. A van, articulated truck and electric urban vehicle are not interchangeable inputs to one undifferentiated truck factor.

Imagine the cargo shares a vehicle with other customers' goods. The vehicle's total fuel use describes the joint operation; the consignment needs an explained allocation. Neither dividing equally by customer count nor assigning the whole journey to this buyer is justified merely because the invoice is available. Volume constraints and the applicable method can affect the loading interpretation.

Ask how empty running and return operations are represented before adding a separate allowance. A factor already incorporating those operations must not receive the same correction twice. Similarly, zero exhaust emissions from battery-electric driving do not establish zero emissions from supplying its electricity.

Road also remains inside many nominal rail, air and sea offers. The first and last road legs may use different vehicles from a direct long-haul service. Their actual collection and delivery distances belong to the alternative, even where they make a theoretically attractive mainline substitution less compelling.

Rail freight: a route-specific alternative

A door-to-door rail offer contains more than a train journey. Road pickup moves the cargo to an origin hub, the hub transfers it to rail, and an arrival hub and road delivery connect the train to Buyer B. Those activities belong to the same service specification; describing only the rail mainline would compare a fragment with a complete road delivery.

A complete door-to-door rail service encloses road pickup, origin hub, rail mainline, arrival hub and road delivery in that order.
The rail mainline is only one component of this illustrative complete offer. Arrows show sequence, not time or emissions shares.
Diagram text description

One enclosure groups the five included operations of the illustrative rail service: road pickup, origin hub, rail mainline, arrival hub and road delivery. The arrows represent cargo progression. Every shown movement and hub needs appropriate emissions coverage without duplicate bundled totals. The drawing is not to scale and does not prescribe every real rail route or prove that a schedule or capacity is available.

The sequence shown here is an illustrative rail arrangement, not a universal terminal design. A private siding or different carrier network can change the operations needed. Match the offered terminal pair, traction energy and schedule to the cargo rather than importing an unrelated national average. Where electricity is used, retain its appropriate energy-provision basis; diesel traction needs its corresponding fuel treatment.

Availability can decide the outcome before intensity does. A rail departure that misses the collection cut-off or lacks capacity cannot deliver the hypothetical 50-hour promise. Confirm transfer windows, final delivery and any in-chain storage as part of the actual offer. A physically coherent route and a low modelled figure are useful, but neither alone proves the service can be booked.

Sea and inland waterways: efficient averages, distinct services

Waterborne freight includes very different services. An ocean mainline vessel, coastal feeder and inland barge have different routes, capacities and operating conditions. Broadly favourable unit intensities can justify exploring a water route; they do not establish the footprint or transit time of the service available to this buyer.

For a particular offer, follow the goods from inland pickup to the departure port, any feeder or transshipment, the main vessel and onward delivery. Do not assume every route uses every stage, but do not stop at the main vessel when another stage is required. Port handling and cargo allocation need compatible evidence. A vessel's entire voyage total is not the emissions share of one container or this 2-tonne consignment.

Service frequency and cut-offs matter alongside sailing time. A ship that takes little time underway can still offer a long collection-to-delivery interval if cargo waits for departure or onward distribution. In the fictional decision later, the lowest emissions estimate fails the common deadline for exactly this reason. That is a service constraint, not proof that waterborne transport is generally unsuitable or that another ship on the same coast would produce the same result.

Calculate the complete multimodal chain

List each movement and hub before summing the chain. The working register should identify provider, cargo mass, leg endpoints, service dates, distance convention, method or factor version and whether another provider figure already covers it. Sum shipment-allocated results on compatible gas and fuel/energy boundaries, not raw tonnes, kilometres or mixed CO2 and CO2e figures.

Continue the invented 2-tonne rail case. All distances, coefficients and allocated hub results below are made up. Assume compatible energy-provision-inclusive CO2e boundaries, with appropriate loading and empty operations already represented in the fictional transport coefficients. They demonstrate arithmetic, not a factor dataset or a claim of compliance.

Chain componentInvented activity and coefficientCalculationContribution
Road pickup2 tonnes; 50 km; 0.10 kg CO2e per tonne-km2 × 50 × 0.1010 kg CO2e
Origin hubAllocated shipment resultNo distance factor applied5 kg CO2e
Rail mainline2 tonnes; 1,000 km; 0.02 kg CO2e per tonne-km2 × 1,000 × 0.0240 kg CO2e
Arrival hubAllocated shipment resultNo distance factor applied7 kg CO2e
Road delivery2 tonnes; 40 km; 0.10 kg CO2e per tonne-km2 × 40 × 0.108 kg CO2e

Table 2, shown one row at a time. Every field is retained.

Row 1

Chain component
Road pickup
Invented activity and coefficient
2 tonnes; 50 km; 0.10 kg CO2e per tonne-km
Calculation
2 × 50 × 0.10
Contribution
10 kg CO2e

Row 2

Chain component
Origin hub
Invented activity and coefficient
Allocated shipment result
Calculation
No distance factor applied
Contribution
5 kg CO2e

Row 3

Chain component
Rail mainline
Invented activity and coefficient
2 tonnes; 1,000 km; 0.02 kg CO2e per tonne-km
Calculation
2 × 1,000 × 0.02
Contribution
40 kg CO2e

Row 4

Chain component
Arrival hub
Invented activity and coefficient
Allocated shipment result
Calculation
No distance factor applied
Contribution
7 kg CO2e

Row 5

Chain component
Road delivery
Invented activity and coefficient
2 tonnes; 40 km; 0.10 kg CO2e per tonne-km
Calculation
2 × 40 × 0.10
Contribution
8 kg CO2e

The chain total is 10 + 5 + 40 + 7 + 8 = 70 kg CO2e for the fictional shipment. Each movement follows tonnes × kilometres × kg CO2e per tonne-kilometre; each hub contribution is already allocated in kg CO2e. A hub is not assigned an imaginary travel distance to make it fit the movement formula.

For example, suppose the rail provider supplies a bundled 52 kg CO2e covering the 40 kg mainline and both hubs, 5 + 7 kg. Use pickup 10 + bundle 52 + delivery 8 = 70 kg CO2e. Do not add that bundle to the five-row total: it replaces its covered components. A genuinely door-to-door 70 kg result would replace all five rows, subject to verifying the stated coverage. Keep unknown or incompatible coverage unresolved rather than adding whichever figures happen to be available.

Factor in lead time and inventory without hiding transport

A mode change can alter when stock must leave the supplier and how much inventory is in transit. Those consequences matter to the business, but a financing cost or additional stock quantity is not itself a transport emissions result. Present delivery time, transport CO2e, inventory consequences and price as distinct measures before deciding whether a broader assessment is needed.

For example, a buyer using a regular slower service might place orders earlier to maintain the same replenishment date. It might instead need extra stock, or be unable to change the ordering cycle. None of these outcomes should be assumed from the mode label alone. Handling and storage already included inside a transport hub's boundary remain there; do not add them again as a separate warehouse penalty. Additional business storage outside that declared freight boundary needs its own activity data and method if assessed. A wider logistics decision can combine compatible quantified consequences, but should still show the transport result separately so that a claimed modal improvement is not hidden inside an unexplained net total.

Rank methods by evidence quality

Prefer representative physical evidence, not simply the smallest number or the largest number of decimal places. Traceable carrier activity and allocation can support a route-specific assessment. A suitable mass-distance model can also be useful, especially when actual service information is incomplete. Spend-only estimates answer a coarser screening question.

Evidence availableUseful decision supportLimitation to resolve
Carrier fuel or energy with shipment allocationExplains the relevant operation and consignment shareVerify period, coverage, allocation and representativeness; carrier branding is not validation
Mode-distance model with applicable intensityCompares defined movement legs on a consistent methodTest vehicle or service class, distance convention, loading and factor version
Spend-based screeningIdentifies purchases worth investigatingPrice changes do not establish changes in physical freight emissions

Table 3, shown one row at a time. Every field is retained.

Row 1

Evidence available
Carrier fuel or energy with shipment allocation
Useful decision support
Explains the relevant operation and consignment share
Limitation to resolve
Verify period, coverage, allocation and representativeness; carrier branding is not validation

Row 2

Evidence available
Mode-distance model with applicable intensity
Useful decision support
Compares defined movement legs on a consistent method
Limitation to resolve
Test vehicle or service class, distance convention, loading and factor version

Row 3

Evidence available
Spend-based screening
Useful decision support
Identifies purchases worth investigating
Limitation to resolve
Price changes do not establish changes in physical freight emissions

Keep the factor publisher, version, geography, applicable year and energy boundary with the result. For an electrified option, confirm how charging or traction electricity is represented rather than replacing it with a tailpipe-only comparison. Disclose exclusions and what remains uncertain. If two totals differ only slightly but loading or route evidence is weak, the apparent ranking may not support a confident procurement claim. Request the missing relevant evidence rather than inventing a precise uncertainty interval or treating every carrier-supplied result as automatically superior.

Test a real lane before making a modal claim

Apply the same fictional task to all four offers: 2 tonnes and 4 cubic metres of the specified boxed parts, Factory A to Buyer B, delivery within 72 hours, with the same handling requirements and compatible CO2e coverage of required movements and hubs. Assume cargo acceptance and capacity are confirmed solely for this teaching scenario. The schedules and complete-service estimates are invented; only the rail total is worked through above. No real route, provider performance or modal saving is asserted.

Fictional offerComplete included serviceAssumed door-to-door resultDecision under the fixed task
AirRoad access, airport handling, flight and final delivery18 hours; 190 kg CO2eMeets 72 hours; investigate whether its extra speed has any required value
Direct roadCollection, complete road movement and delivery30 hours; 110 kg CO2eMeets 72 hours; retain as the reference offer
Rail chainPickup, origin hub, rail, arrival hub and delivery50 hours; 70 kg CO2eMeets 72 hours; lower fictional estimate than road, conditional on matched evidence
Coastal sea chainPickup, ports, sailing and final delivery90 hours; 55 kg CO2eFails 72 hours; not an equivalent replacement despite the lower figure

Table 4, shown one row at a time. Every field is retained.

Row 1

Fictional offer
Air
Complete included service
Road access, airport handling, flight and final delivery
Assumed door-to-door result
18 hours; 190 kg CO2e
Decision under the fixed task
Meets 72 hours; investigate whether its extra speed has any required value

Row 2

Fictional offer
Direct road
Complete included service
Collection, complete road movement and delivery
Assumed door-to-door result
30 hours; 110 kg CO2e
Decision under the fixed task
Meets 72 hours; retain as the reference offer

Row 3

Fictional offer
Rail chain
Complete included service
Pickup, origin hub, rail, arrival hub and delivery
Assumed door-to-door result
50 hours; 70 kg CO2e
Decision under the fixed task
Meets 72 hours; lower fictional estimate than road, conditional on matched evidence

Row 4

Fictional offer
Coastal sea chain
Complete included service
Pickup, ports, sailing and final delivery
Assumed door-to-door result
90 hours; 55 kg CO2e
Decision under the fixed task
Fails 72 hours; not an equivalent replacement despite the lower figure

On these assumptions, the sea offer cannot win the specified delivery by being cheapest in carbon. Among feasible offers, rail is the candidate with the lowest fictional transport estimate, not a universal winner. If the buyer's rule is to prefer the lower transport estimate subject to the delivery requirement, rail merits the next procurement check. Actual price, capacity commitments, schedule reliability and any relevant inventory assessment must still be obtained before booking.

The comparison also identifies useful challenges. Ask the air provider whether a less urgent service still meets the deadline. Confirm that road and rail include all required handling on the same basis, and test whether uncertainty could overturn a material claimed difference. A revised sea schedule within 72 hours would be a changed offer to recalculate, not permission to erase the original failure. The same discipline applies if the customer changes the deadline.

Keep a lane-decision record containing the fixed task, candidate offers, included legs, methods, assumptions, evidence limitations, selection owner and update trigger. A different carrier route, loading allocation or promised arrival can require a new comparison. The conclusion supported here is about this defined fictional service; a real lane assessment must replace the invented inputs before it can support a shipment claim.

Carry the service decision into the shipment record

A useful freight comparison ends with a service decision that can be traced to the goods actually being moved. Product identifiers and shipment evidence serve different jobs: an SKU or bill of materials identifies an item, while consignment mass, route, carrier activity and allocation establish the transport calculation. Matching an identifier is not proof that those transport inputs are complete.

For a practical handoff, connect the selected offer to the actual shipment reference and the relevant product record. Reconcile the shipped quantity and packaging rather than assuming a catalogue mass describes the dispatched load. Keep the transport method and its reviewer identifiable. A product-data record does not by itself provide carrier integration, modal optimisation or an independently validated emissions engine.

The decisive question remains whether comparable evidence supports a feasible delivery. Keep cargo, endpoints and the service requirement stable, include the necessary connectors, and make unresolved differences visible. Then choose or investigate an offer under the buyer's stated priorities. Do not extend one lane's conditional result into a promise about every journey made by that mode.